1. INTRODUCTION
Towards the end of the last century and in the beginning of this century many papers on desmids in some part of Scandinavia have been published: Grönblad (1920, '21, '34, '42, '46, '47, '48), Boldt (1887), Borge (1892, 1913), Nordstedt (1873), Wille (1881). The study of desmids was a rather new branch of science in that time, and most research was directed towards describing the large number of species. Therefore the articles usually had a taxonomic and/or floristic character. Although these papers often included a description of the environment, there was never paid much attention to the relations between this environment and the desmidflora. Strøm (1926) was one of the first authors who described the ecology of some Scandinavian desmids. Flensburg (1967) described the relations between desmids and trophic status, pH and conductivity. In Flensburg and Malmer (1970) the relations with the macrophytic vegetation was added to these factors
After the rise of the phytosociological schools in the twenties and thirties it became popular to describe vegetation in terms of communities. When this system was applied to "Desmid vegetations" it appeared that it was not so easy to recognize clear communities. Even when the surrounding macrophytes showed no difference, the composition of the Desmid vegetations could vary considerably. According to Heimans (1937, '54) this was due to an inadequate accessibility, which means that an environment that is suitable for a species to live in, not necessarily needed to contain this species, because this species simply did not reach this place (yet). Coesel (1981) mentions the short life cycle of desmids and the easy displacement of individuals from one place to another as a cause for the difficult delimitation of desmid communities. In spite of all these problems it may still be useful to try to find some system in the distribution of desmids. Coesel (l.c): "since,in the view of the knowledge that desmids have such narrow and usually diverse ecological amplitudes, there is still a demand for a typology of desmid assemblies, by means of which, for instance, small scale environmental changes can be detected". Coesel (l.c.) made such a typology of desmid assemblies in a Dutch broads area, with as main habitats shallow peat pits and quivering fens.
As there is still very little known about desmid communities, I intended to do some further studies, especially with concern to the oligo‑ to mesotrophic benthic vegetations. The investigation area is the Skogvoll mire reserve at Andøya, an isle in northern Norway. Andøya is for a large part covered with peat, part of which is protected in the "Skogvoll mire reverve". The reserve is a very suitable place for investigating desmids.Scattered in the mire there are thousands of waters, comprising small pools, small to bigger lakes, soaks, springs and streams, with a trophic level varying from oligotrophic to eutrophic. Due to this variation it is an ideal environment for studying the effect of various ecological factors on the composition of desmid communities. The suitability of this place is reflected in the large amount of species that was encountered. Within the actual sampling area, that covers about 30 km2 of different types of (oceanic) mire, over 450 desmid species have been recorded.
The purpose of this study was to reveal the relations between on one hand desmid species and desmid communities, and on the other hand their environment, represented as vegetation, chemical parameters, and other observations made in the field.
Classification was done by means of computer processing. It was possible to delimit 13 clusters of desmid species, which could be correlated with relevant ecological parameters.